ChanServ changed the topic of #linux-sunxi to: Allwinner/sunxi development discussion - Don't ask to ask. Just ask! - See http://linux-sunxi.org | https://github.com/linux-sunxi/ | Logs at http://irclog.whitequark.org/linux-sunxi
chujalt has quit [Quit: Saliendo]
_whitelogger has joined #linux-sunxi
anunnaki has joined #linux-sunxi
egbert has quit [Disconnected by services]
egbert has joined #linux-sunxi
theOzzieRat has quit []
theOzzieRat has joined #linux-sunxi
rz2k has joined #linux-sunxi
ZaEarl_ has joined #linux-sunxi
ZaEarl has joined #linux-sunxi
ZaEarl_ has quit [Read error: Connection reset by peer]
hramrach has quit [Remote host closed the connection]
wingrime has joined #linux-sunxi
ZaEarl_ has joined #linux-sunxi
ZaEarl has quit [Read error: Connection reset by peer]
ZaEarl_ has quit [Ping timeout: 240 seconds]
rellla has joined #linux-sunxi
rellla2 has joined #linux-sunxi
rellla has quit [Ping timeout: 260 seconds]
wingrime has quit [Ping timeout: 260 seconds]
rz2k has quit []
* mnemoc wonders what's the pitch of that monster 2x35 (unpopulated) header in the marsboard
<rm> looks like the "standard" pitch :)
<rm> like on the Raspberry Pi and Arduino and thousands of devices e.g. jumpers on ATX motherboards
<rm> I don't remember how much in mm
<mnemoc> the usual is 0.1" (2.54mm)
<mnemoc> the cubieboard uses 2mm
<rm> ok, so it looks to be 2.54mm
<mnemoc> but this thing looks too small even for that
<rm> hm, or maybe not
<rm> you can measure it compared to the sata socket
<rm> in pixels
<rm> or even the A10
<mnemoc> yes, you are right. it seems 0.1" in that picture
<mnemoc> nope. 2mm
<rm> I don't like they are unpopulated
<rm> at all
<mnemoc> and not visible uart pads
<rm> maybe UART is somewhere among those headers
<rm> looks like way too many pins
<mnemoc> true
<mnemoc> weird thing they mention a wifi chip in the specs, but it's not on the pictures
<rm> specs are copypasted
<rm> as someone noted on the MLs
<mnemoc> no PMU either
hansg has joined #linux-sunxi
<mnemoc> really hope hipboi can start producing the cubieboard in decent quantity soon
<mnemoc> he is wasting his momentum
<oliv3r> for the A20? (sorry just jumping into the last 2 lines of this convesation, reading back now!)
<mnemoc> even if it's the same A10
<rm> afaik his plan was to switch to A20 asap
* mnemoc desparately wants an A20 (and possibly A31) device with uart
zenitraM has joined #linux-sunxi
rellla2 is now known as rellla
<mnemoc> i'm adding the pinout I got for the A10 SO-DIMM from "bdd", there are two pads called CSI0-IO-2V8 and CSI1-IO-2V8... what's that?
<mnemoc> never seen "CSI" before
<RaYmAn> CSI is usually camera
<RaYmAn> MIPI-CSI
<RaYmAn> well, it stands for Camera Serial Interface, so I guess it is :P
<mnemoc> ah, cool. thanks :)
<mnemoc> it was colored in the schematics as coming from the AXP so I got confused
<RaYmAn> well, makes sense if it's the power line
paulk-desktop has joined #linux-sunxi
<rm> mnemoc, "2.0mm Spaced GPIO Headers", from http://www.hotmcu.com/marsboard-a10-dev-board-p-59.html?cPath=33
<gzamboni> mnemoc its the 2.8V required for the camera sensor
<mnemoc> gzamboni: those those have some logic or just dumb VCC 2V8 ?
<mnemoc> rm: :)
<gzamboni> just dump vcc2v8
<gzamboni> its named like that as the csi is the only place that need this voltage
<gzamboni> you gotta check if the camera sensor you will use needs 1.8V also
<gzamboni> there are some camera sensors with an internal regulator and others without
<gzamboni> my pcb is in the revision state, i think tomorow it will be done
<gzamboni> if you want i can send you the schematics in about 4 hours as they are not here in this pc
hipboi has joined #linux-sunxi
<gzamboni> nice board rm
<gzamboni> didnt know this one
<gzamboni> hi hipboi, can i /q ?
<hipboi> gzamboni: hi
<mnemoc> gzamboni: that would be great. thanks. and thanks for the info. I'll adjust the wiki page accordingly.
<mnemoc> wb hipboi!
<hipboi> hi
<hipboi> the last week we are busy with moving and declorating the office
<mnemoc> nice :) in zhuhai?
<hipboi> yes
<oliv3r> hipboi: Cubieboard A20! Any idea's yet? :)
<gzamboni> ok mnemoc
hipboi has quit [Remote host closed the connection]
<oliv3r> fine! leave!
hipboi has joined #linux-sunxi
<gzamboni> he answered you
<gzamboni> with acts
<oliv3r> lol
<hipboi> A20 is taped out now
<oliv3r> so hipboi, any idea/eta/info on CubieBoard A20
<hipboi> allwinner is debugging the chip
<oliv3r> oh nice, that's good right?
<hipboi> not sure when they will release the source code
<oliv3r> I saw some software support dumps here and there, so it's going well?
<oliv3r> are they planning on it?
<hipboi> isn't there is a mysterious guy leaked the source code?
<oliv3r> I guess I should have said, is allwinner planning on releasing it officially?
<oliv3r> tbh, they should just put their git public :p, or atleast push stuff occasionally there when it's release time :p
<hipboi> they would not release publicly
<oliv3r> so they leak instead :p I guess that's 'just as good' as complieing in this sense
<oliv3r> hipboi: I have heard that allwinner has sanctioned that the A13 datasheet was allowed to be shared publically
<oliv3r> (it's identical to the A10 one we allready have, the one from the ML, with sections masked of course)
<oliv3r> hipboi: any plans on the A20 datasheet?
<oliv3r> or do we have to hope for a leak there aswell?
<hipboi> hope for a leak
<hipboi> leaking is publishing....
<hipboi> in some way
<hipboi> mnemoc: where did you push the leaked a31 source code?
<mnemoc> on my github
<hipboi> ok, i see it
<mnemoc> it's from the first leak, without history
<mnemoc> s/leak/source drop/
<hipboi> there is tag
<oliv3r> hipboi: cloudsto announced a few A20 'to be expected mid-march' how likly do you think that will be? I guess cubeboard 2 won't make it by that time?
<hipboi> mid-march?
<mnemoc> hipboi: from a10/a13/a10s sources, those are in the lichee/ branches of my github
<mnemoc> still need to make a sanitization branch, preserving the history, but fixing encoding, ^M and whitespacing
<gzamboni> anyone have/know the pinout of this marsboard gpio ?
<mnemoc> lichee/ branches are pristine
<mnemoc> gzamboni: mail them :)
<gzamboni> will do so
<hipboi> reuuimllatech.com
<hipboi> i named it before i left allwinner
<oliv3r> I'm so excited for those branches. Intentionally or not, it deserves a padding on their shoulder for leaking it :)
<mnemoc> hipboi: was the reuuimlla word invented by you or your boss?
<hipboi> me
<mnemoc> it's nice :)
<hipboi> i replaced all allwinner to reuuimlla
<jinzo> but the domain isn't even registered? Could be quite a fun phising target.
<hipboi> reverse of allwinner
<hipboi> come on, it's a fake name
<mnemoc> there is a newbietech too
<hipboi> yes
<hipboi> another fake name
<mnemoc> https://github.com/amery/git-import-help/blob/master/reuuimlla.sh is part of the sanitization routine :p
<jinzo> hipboi, yeah I know I know. But still. If someone would to register the name and go on how they're the new allwinner and whatnot
<hipboi> do we need to make a board with a31?
<hipboi> or better choice like imx q
<jinzo> A20 would probably be more desirable around here, but I would guess noone would object if you make A31
<mnemoc> hipboi: the imx q market already has several boards. A20 and A31 has none
<jinzo> There're quite some boards with i.MX Q - but I have yet to see any of A31
<jinzo> or what mnemoc said ^^
<hipboi> hmm...
<mnemoc> stick with allwinner chips, specially now that we have sources
<gzamboni> got the schematics of the marsboard
<mnemoc> and you'll sell all the stock you can manufacture
<mnemoc> gzamboni: wiki?
<gzamboni> yes, one sec
<jinzo> I would agree with mnemoc - and I hope you have the capability to design and manufacture allwinner based devices in a timely manner
<jinzo> I see that cubieboard is in stock now
<jinzo> or at least semi available
<jinzo> er, I was mislead, still hard to get by
<hipboi> mnemoc: where is the original link you download the source code?
<mnemoc> it wasn't public
<mnemoc> gzamboni: thanks!
<gzamboni> np
<hipboi> mnemoc: it wasn't public?
<mnemoc> hipboi: got the link to some random machine in amazon's cloud service in private
<hipboi> hmm...
<mnemoc> tarballs of lichee things made with repo from a 172.something git server
<mnemoc> but symlinks were textified so I guess they were tared on a windows machine
<hipboi> 172. is the internal server of allwinner
<hipboi> so it's intended leaking
<oliv3r> very well done :)
<oliv3r> now, if the yleak the cedarx sources, I will send them a birthday present :)
<mnemoc> :D
<oliv3r> Big cheese (or who was it again that liked cheese?)
<hipboi> :)
<oliv3r> "to the Allwinner dev team"
<oliv3r> i don't think a cake would survive the postal route :(
<mnemoc> afaik .cn people isn't very fond to sugar
<oliv3r> So a big cheese would be good? Gouda? :)
<oliv3r> Sweet and sour sauce is sugary!
ganbold_ has joined #linux-sunxi
rz2k has joined #linux-sunxi
ganbold_ has quit [Remote host closed the connection]
<oliv3r> can we easily extract a script.bin from a livesuit image file (or a fex file)?
<mnemoc> there are open source packing/unpacking tools
<oliv3r> i'll try them on my 'original' image then and see what i get out of it
<oliv3r> i guess if I search the wiki for awutils i'll find interesting stuff
<mnemoc> hope so
<oliv3r> nothing
<hipboi> there is a window tool
<hipboi> called unimg
<hipboi> to extract the livesuite image
<hipboi> written by a Chinese boy
<hipboi> at the time of F20
<mnemoc> sure, but I prefer to promote the open tool ;-)
ganbold_ has joined #linux-sunxi
<oliv3r> who is this ithmar anyway and why did he write that?
<mnemoc> a gut from .nl
<mnemoc> guy*
<mnemoc> libv knows him
<libv> he did telechips stuff a few years ago
<libv> the only one touching telechips tablets with a clue
<oliv3r> ah, hup holland hup :p
<oliv3r> i was just supprised as i'd imagine it falling somewher under linux-sunxi, a wiki page or him idleing here
<mnemoc> hipboi: who ever is in charge of drivers/usb/serial/option.c in allwinner's kernel. I hate him
<hipboi> me too
<mnemoc> *g*
hipboi has quit [Quit: Leaving]
FergusL has left #linux-sunxi ["Quit"]
<oliv3r> sweet, that works, it decompressed the image and inded the script.bin (seems uncompressed) is in there
<oliv3r> seems in fex format
<arete74> oliv3r:was is the name the file "script.bin" after decompress image?
<oliv3r> 12345678_1234567890script
<oliv3r> 2014 bytes
<oliv3r> speaking of, do we have a 'most basic script.bin' that we could use to base our device uppon? I mean, we know that script.bin just gets copied and some minor modifications done to it
<oliv3r> so a bare minimum to 'start from' would be nice, no?
<oliv3r> ZaEarl, try using gcc-4.6.3 from linaro
<rz2k> anyone designed boards with a13?
<oliv3r> rz2k: anyone here? or generally?
<rz2k> cant find ball-to-die info for ddr3 routing
<rz2k> here
<mnemoc> rz2k: Tsvetan must have the reference design
<rz2k> cant ask olimex for doing support instead of allwinner
<rz2k> dont ask why :p
<mnemoc> but you can ask for the pdf
<mnemoc> it might even be in olimex's github
ganbold has quit [Ping timeout: 255 seconds]
ZaEarl_ has joined #linux-sunxi
gzamboni has quit [Read error: Connection reset by peer]
gzamboni has joined #linux-sunxi
wingrime has joined #linux-sunxi
rellla has quit [Ping timeout: 272 seconds]
hansg has quit [Quit: Leaving]
rellla has joined #linux-sunxi
<wingrime> mnemoc: when you will make patch review ?
<mnemoc> wingrime: what patch exactly? pretty busy with work and cleaning/rebasing the history of the A10/A13/A10s/A20/A31 code I got
<wingrime> [PATCH] ft5x_ts:enable power port hardware featue
<wingrime> this is a last patch for support ft5x_ts for my hw
<mnemoc> wingrime: will need an ack from drachensun and techn for that
rellla has quit [Quit: Nettalk6 - www.ntalk.de]
<wingrime> mnemoc: ok
hramrach has joined #linux-sunxi
<techn__> mnemoc: wingrime: I gave comments about coding convension
<techn__> wingrime: asked if he can do patch afterwards for whole driver :/
<wingrime> techn__: I followed driver code style, if you wanna make do driver cleanup for code style I think it better do in separate patch afterwards
<wingrime> techin__: I can make some driver cleanup
<techn__> wingrime: ft5x_set_reg(FT5X0X_REG_PMODE, PMODE_HIBERNATE); , should that be done on suspend.. if it makes controller do some power save stuff?
<wingrime> techin__: if HW have power port feature I simply poweroff TS IC
<techn__> or will pin control shutdown whole chip?
<wingrime> yes, thats why it not worked when I tested it first time
<wingrime> i do some IDA investegation and saw this stuff
<wingrime> if this pin set to 0 chip not answer over I2C
<wingrime> just timeout messages
<wingrime> also there was IRQ problem , but it already fixed
<wingrime> techn__: so what should I do ?
<techn__> wingrime: I found other improvement.. I'll send it to ML
<mnemoc> wingrime: about style, in this context it's prefered to fix style on the go. new code must have correct linux style, and only lines affected by code changes get fixed
<mnemoc> and allwinner doesn't have a code style at all. they don't even care to align the lines within the same if () { }
<wingrime> techn__: ok
<wingrime> if I send fixed patch will it be attached to same thread on ML?
<techn__> wingrime: that's good question.. dunno the practice :/
<mnemoc> wingrime: git-send-email supports arguments for that
<mnemoc> you give the id of the mail you are replying, and the new mail is composed accordingly
<wingrime> example please
<techn__> mnemoc: is that prefferred way.. Even though I have used that reply-to feature
<techn__> ?
<wingrime> and where can I get this "id"?
<mnemoc> techn__: it's supposed to work.... haven't tried myself
<mnemoc> wingrime: in the header of the mail
<mnemoc> Message-Id: <1362432598-14569-1-git-send-email-wingrime@gmail.com>
<techn__> In gmail "show full message"
<mnemoc> meh... wanted to cut the tail :<
<wingrime> and what command line argument for it?
<mnemoc> man git-send-email ? :)
<wingrime> I just think you remeber it....
<mnemoc> never used it myself
<mnemoc> just theory
<mnemoc> i don't even know if it works
<wingrime> more one qusetion it depens use hardfloat toolchain or simple eabi for u-boot ?
<mnemoc> mripard: btw, miniand did not manufacture/design the mini-x. it's made by PineRiver. miniand is just a rebrander
rz2k has quit []
<mnemoc> wingrime: u-boot is float-free, both should work
<wingrime> my a13 tablet have zet6221_ts
<wingrime> we have source for this one ?
<mnemoc> wingrime: no
<wingrime> techn: what difference between wakeup and resume ? one more early than other ?
<techn__> resume calls wakeup
<techn__> wingrime: ^
<techn__> ctp_ops.ts_wakeup();
<wingrime> techin; sorry just noticed
rellla has joined #linux-sunxi
<wingrime> techn: resume also do reset
<wingrime> it I move pin logic to wakeup, it do reset then powerup
<wingrime> *if
<wingrime> useless make reset before power enabled
<wingrime> I don't wanna touch this logic becose power powrt feature are optional
<wingrime> current my logic: if power port used than power on device and wait init 48 ms otherwice send reset and wake up.
<hramrach> hello
<hramrach> anybody knows why is the sd casrd size limited to 32G when sdxc is supported?
<mnemoc> sdxc is supported where?
<hramrach> 32G is sdhc max, right?
<hramrach> there is some code for it
<hramrach> maybe it's just dead code, don't know
<drachensun-peng> I have tested a 64GB card, it worked on mine
<mnemoc> uh
<drachensun-peng> I guess I didn't pull a full byte by byte check
<hramrach> CC drivers/mmc/sunxi-host/sdxc.o
<hramrach> also the whining about lack of sdxc on sdhc cards when you plug them in
Guest62243 is now known as fredy
<drachensun-peng> Mine says "mmc0: new high speed SDXC card at address e624"
<drachensun-peng> everything looks good
<mnemoc> a10?
<drachensun-peng> yeah
<drachensun-peng> now this is one of my with Linaro on the internal flash
<drachensun-peng> but I just tested both SD card slots and they both seemed to work
<drachensun-peng> so I don't have the community U-boot in this case, I am using the Android U-boot and boot0/1 to start it up
<drachensun-peng> but its the stable 3.0.57 kernel from about a month ago
<drachensun-peng> hramrach what are you on?
<mnemoc> if you want to run 3.4 you'll need to replace that stock u-boot with one built from our lichee-dev branch
<drachensun-peng> yeah, I've been meaning to get back and figure that out
<drachensun-peng> does the lichee-dev branch have spl also or do you have to leave boot0/1?
<mnemoc> relies in boot0/boot1
<drachensun-peng> ok, I kind of like that because the screen will be initialized then
<mnemoc> it's basically the same as the stock u-boot.bin, but fixed to pass the machine id and mem info correctly
<wingrime> afull.... a13 have strange wi-fi IC rtl8188e ....
<techn__> wingrime: ok.. so power can be specified togeter with reset and wakeup pins.. then your implementation makes sense
rellla has quit [Quit: Nettalk6 - www.ntalk.de]
<wingrime> techn__: I not very sure that reset is mandatory after hibernate but this code actualy not mine, i think that wake up is only one have to be done but it need invetegate
<wingrime> it need be connected with fex infrastructure
<wingrime> mnemoc: what do you think about it ?
<mnemoc> sure. fix the style issues, adapt it for script.bin usage and submit it to the ML
wingrime has quit [Ping timeout: 272 seconds]
<hramrach> drachensun-peng: a10 too
<hramrach> only sdhc card
paulk-desktop has quit [Read error: No route to host]
chujalt has joined #linux-sunxi
<chujalt> Hola, saludos
ganbold_ has quit [Remote host closed the connection]
<hramrach> anyone familiar with block layer?
<hramrach> I just found that I have block devices that are present but somehow inactive
<hramrach> eg. the loop devices and initrds
chujalt has quit [Quit: Saliendo]
<hramrach> hmm, loop just creates a number of devices
<hramrach> nothing to do about that except not loading loop if you don't want the mess
<hramrach> they aren't actually in any way inactive, just zero size
ganbold has joined #linux-sunxi